Digitalization of 3-D Human Bodies: A Survey
Yanhong Yang, Haozheng Zhang, Alfredo Ballester, Sandra Alemany, Shengyong Chen, Guodao Zhang
Abstract
Human plays a key role in metaverse and digital twin systems and thus live three-dimensional (3D) human shape reconstruction brings significant improvement for highly accurate anthropomorphic measurements. However, it is still difficult to obtain detailed 3D computational models or “virtual humans” due to the diversity of individual poses and shape deformation. With technical advantages of the metaverse geometric modeling, virtual characters can be created in seconds. Concurrent, tremendous progress has been achieved in 3D modeling technology and larger representative public human datasets have emerged. In this survey, the status quo of 3D body digitization is investigated, and the accompanying processing pipeline is introduced. Specifically, this survey first discusses mainstream 3D scanning technologies for virtual 3D body acquisition. Second, we investigate the status quo of the commonly-used body measurement processing frameworks from four aspects: 1) data acquisition; 2) feature extraction; 3) surface modeling; and 4) landmarking and digital measurements. Finally, through a comparative analysis of prior works, we highlight challenges and new trends of ongoing developments based on insights into the advantages and limitations of existing works.